Ni Stir 5591

NISTIR 5591
Post-Occupancy Evaluation
of the Forrestal Building
Philip A. Sanders
Belinda L. Collins
U.S. DEPARTMENT OF COMMERCE

Technology Administration
National Institute of Standards
and Technology
Building and Fire Research Laboratory
Gaithersburg, MD 20899
Prepared for:
U.S. Department of Energy

Washington, D.C.
NIST
Post-Occupancy Evaluation
of the Forrestal Building
Philip A. Sanders
Belinda L. Collins

Technology Administration
National Institute of Standards
and Technology
Building and Fire Research Laboratory
Gaithersburg, MD 20899
Prepared for:

Washington, D.C.
March 1995
★ ★

Ronald H. Brown, Secretary
TECHNOLOGY ADMINISTRATION
Mary L. Good, Under Secretary for Technology
NATIONAL INSTITUTE OF STANDARDS

AND TECHNOLOGY
Arati Prabhakar, Director
ABSTRACT
A post-occupancy evaluation was performed on the Department of Energy Headquarters Building

(the Forrestal Building) in Washington, D.C. The lighting in the building was retrofitted with
new, more energy-efficient, components to meet energy target guidelines. Occupant responses
to the indoor environmental conditions, particularly the lighting, were studied to determine the
impact of the relighting on the building inhabitants. In addition, physical measures of the
lighting and other environmental conditions before and after the relighting were compared. The
post-occupancy evaluation employed a questionnaire about the environmental conditions and
physical measures of the space (lighting, space, noise, temperature, etc.). A total of 244/220
people participated (before and after the relighting, respectively). Physical measures were taken
at 100 work stations before the relighting and 75 after. Analysis of the physical measurement
data indicated generally higher lighting levels with more even distribution of luminances in the
offices. Occupant response to the changes in the lighting was generally quite positive. The
relighting was perceived to have improved the appearance of the building substantially, as well
as the lighting within individual workstations. Finally, guidance is given for doing post¬
occupancy evaluations as part of other relighting initiatives.
KEYWORDS:
building technology, contrast, energy efficiency, environmental assessment, federal relighting,
illuminance, lighting, luminance, post-occupancy evaluation, temperature, VDT’s.
FOREWORD
This report summarizes research conducted from January 1993 to December 1994 under contract
Number DE-AI01-94CE73288.
ACKNOWLEDGEMENTS
The authors wish to thank Lou Harris, Jim Vejda and Mike Shincovich of the U.S Department
of Energy for their co-operation in arranging access to the building to collect the data and for
working with us to develop the plans for data collection. The authors also wish to thank Steve
Treado, NIST, for his insight on developing the automatic data collection device. In addition,
the authors wish to thank Andy Persily and Wen Lu of the NIST Indoor Air Quality Group for
their cooperation in collecting the physical measurement data. This report would not have been
possible without their willing and enthusiastic help and support.
DISCLAIMER
Any commercial products or trade names mentioned in this report are included for informational
purposes only, and do not constitute an endorsement or recommendation by the National Institute
of Standards and Technology or the U.S. Department of Energy.
IV
Table of Contents
1. Introduction. 1
1.1 Background. 1
2. Technical Approach. 3
2.1 Background. 3
2.2.1 Questionnaire Survey . 4
2.2.2 Physical Measurements . 4
3. Questionnaire Results. 9
3.1 General. 9
3.2 Subjective Response to Lighting . 10
3.3 Thermal Environment. 16
3.4 General Appearance of the Offices. 16
3.5 Reasons for Choices . 19
4. Measurement Results . 30
4.1 Lighting Measurements. 30
4.2 Other Environmental Measures. 39
5. Conclusions. 41
6. Recommendations for Completing Further Post Occupancy

Evaluations . 42
6.1 Determining the Sample Size and Composition . 42
6.2 Suggested Questionnaire Contents . 44
6.3 Suggested Physical Measurements . 45
6.4 Collecting and Analyzing the Data. 45
Appendix A. Suggested Questionnaire Items. 49
Appendix B. Suggested Physical Measurement Form . 55
v
List of Figures
Figure 1. Ratings of the Quantity of Light as a Function of Sequence. 11

Figure 2. Ratings of the General Appearance of the Lighting System as a
Function of Sequence. 13
Figure 3. Ratings of Control of Light Sources as a Function of Window
Orientation . 14
Figure 4. Ratings of the General Amount of Space as a Function of Window
Orientation . 15
Figure 5. Ratings of the Thermal Environment as a Function of Window
Orientation . 17
Figure 6. Ratings of the Overall Appearance of the Office as a Function of
Windows. 18
Figure 7. Ratings of the Noisiness of the Space as a Function of
Orientation . 20
Figure 8. Percentage of Respondents for Choices . 21
Figure 9. Distribution of Measured Illuminances at the Primary Work
Station. 34
Figure 10. Mean Horizontal and Vertical Illuminance as a Function of
Sequence. 35
Figure 11. Pattern of Lights and Luminance as a Function of Sequence. 37
Figure 12. Shifts in CIE x,y Coordinates for Color Samples as a Function of
Sequence. 38
Figure 13. Measured Temperature and Humidity Data . 40
List of Tables
Table 1. Summary of Questionnaire Results for Sequence and Window

Orientation . 22
Table 2. Summary of Physical Measurement Data . 31
vi
1. Introduction
1.1 Background
Post-occupancy evaluation techniques provide a means for evaluating occupant responses to

changes in an environment and linking this response to physical measures of that environment.
Typically, post occupancy evaluations use a battery of tests to assess environmental conditions
in the facility, including questionnaire surveys of the occupants, physical measures, personal
observations, and individual interviews. The post-occupancy evaluation technique is thus
designed to provide information about the occupants’ reaction to their work spaces and document
the physical conditions to which they are responding - usually on a pre- and post-retrofit basis.
Dillon and Vischer (1987a, b) used post-occupancy evaluation techniques to study four office
buildings in Canada and develop response norms. Rubin and Collins (1987, 1988) and Collins
and Rubin (1988) also used these techniques to evaluate environmental conditions in three U.S.
Army field stations, while Marans (1987), Marans and Brown (1987), Gillette (1988), Gillette
and Brown (1986), and Collins, Fisher, Gillette and Marans (1989) used them in a study of
lighting, energy use, and other environmental conditions in thirteen office buildings in the United
States. Each of these studies identified the importance of lighting, thermal comfort, indoor air
quality and privacy as major factors in influencing human response to an environment.
Collins, Fisher, Gillette, and Marans (1990), Collins and Rubin (1988), Collins, Gillette, Dahir
and Goodin (1989) presented data from post-occupancy evaluations of about 15 facilities in both
government and private industry. These assessments provided information on physical
conditions, especially lighting, and occupant reaction in the various buildings studied. Collins,
et al, (1990) also found noticeable dissatisfaction with some lighting systems, particularly where
fixed task lighting was combined with an indirect ambient lighting system. This particular
system was associated with higher energy use (in terms of lighting power density) and higher
illuminances, but lower ratings of lighting satisfaction than a comparable situation in which
direct ambient lighting provided the primary illumination. Data from the post-occupancy
evaluation of several U.S. Army facilities suggested that lighting levels were below IES
minimum recommendations for the types of task performed in the facilities (Collins and Rubin,
1988; Collins et al, 1989). These two studies also found that occupant satisfaction was below
that obtained in other areas in the same facility, probably because of the lower illuminances.
Data from the Army facilities were used'to suggest areas where lighting retrofits could be
effective, as well as the types of office situations that could be improved.
Other studies have used only questionnaire or laboratory techniques to determine people’s
response to the environment. For example, a telephone survey by Louis Flarris and Associates
(Steelcase, 1987) questioned workers in the U.S. and Canada and found that respondents placed
a great deal of emphasis on office layout, furniture, improved lighting, and chairs as the key to
increasing productivity. These same respondents also placed considerable importance on
1
privacy, improved temperatures, and reduced distractions or noise, as well as reduced glare on
video display terminals (VDT) screens.
Laboratory studies of the response to lighted environments have identified lighting distribution
as a key factor in determining occupant response to a space. For example, Flynn, Spencer,
Martyniuk, and Hendrick (1973) found that ratings of perceptual clarity were dependent on the
overall brightness of the space, while ratings of pleasantness depended on whether the peripheral
room surfaces were lighted and whether the light was distributed uniformly throughout the space.
Lighting installations identified as pleasant typically consisted of peripheral and non-uniform
lighting, while spaces considered to appear spacious were lighted with a bright uniform
distribution of light. Hawkes, Loe, and Rowlands (1979) determined that subjective brightness
and interest were important factors in influencing the evaluation of lighting systems. In addition,
the pattern of luminance was important: designs with only one wall lit were judged as dim, and
designs using only diffuse sources were rated as less interesting than situations which used more
focused sources. Hawkes, et al, suggested that situations that are judged as being brighter and
more interesting (or complex) are also preferred. Recently, when Collins (1993) reviewed the
literature on the psychological response to lighting, she found that, while enabling people to
perform visual tasks, lighting also creates important perceptions, such as pleasantness,
spaciousness, gloominess, and colorfulness. Finally, in a study of hospitals, Ulrich (1987)
determined that windows provide important benefits, including speeding the recovery from
illness to people in confined spaces, such as hospitals and even offices.
Not all environmental assessments have focused on lighting, however. Other studies have
evaluated such environmental issues as thermal comfort, including temperature and humidity.
For example the ASHRAE Standard (55-1981) states that: "80% of all adults dressed for winter
indoor conditions find temperatures acceptable between 68°F and 74.5°F (20-23.6°C), a relative
humidity of 30-60%, and the air velocity at 0.15 to 0.25 m/sec. Acceptable summer indoor
temperature is between 73 and 79°F (22.8 to 26.1°C)." Meyer (1983, p.27) pointed out the
"extensive experimentation has shown that for an average, sedentary, lightly clothed person this
[thermal comfort] occurs most readily when the air in a standard room has a temperature of 24.5
°C (76.1°F), a relative humidity of 40%, and an air velocity of 0.25 m/sec." The IESNA
provides recommendations for minimum illuminance levels, similar in concept to those that
ASHRAE provides for thermal conditions. Thus, the IESNA (1993) suggests that a minimum
of 500 lux is generally advisable for moderate to difficult visual tasks, although their
recommendations are weighted by the age of the observer and the difficulty of the task itself.
These recommendations do not address all the variables responsible for human response to a
lighting environment, however, making it difficult to use a single number as a guide to the
effectiveness of a particular lighting situation.
2
2. Technical Approach
2.1 Background
As part of the Federal Relighting Initiative, the DOE Headquarters in Washington, D.C., was
relighted during the summer of 1993 in a major effort to reduce the electricity consumption of
the lighting system. Indeed, by retrofitting the lighting system in the Forrestal Building, DOE
expects to reduce the electricity consumption of the lighting system by at least 60 percent. The
projected energy savings are the result of retrofitting the old lighting system with new
components. Prior to the retrofit, a portion of the luminaires in each office had been de-lamped
as an energy savings measure creating an irregular and visually distracting pattern of energized
luminaires. During the retrofit, the two T12 cool white fluorescent lamps (nominally 34-40
watts) and magnetic ballasts initially housed in each 30 cm x 120 cm (1 ft. x 4 ft.) prismatic
lensed luminaire were replaced by one T8 lamp and electronic ballasts. In addition, specular
reflectors were added to increase the luminaire’s optical efficiency. Although all luminaires
were energized after the retrofit, the number of energized lamps per office actually decreased.
Collins, et al.(1990), Collins and Rubin (1988), Collins, et al.(1989) recently reported that
employees were noticeably dissatisfied with the lighting in government facilities, especially
where to save energy, lighting levels had been reduced below the minimum IES recommended
illuminances. They concluded that, when lighting systems are retrofitted to conserve energy,
it is critical to maintain, if not improve, the quality of the lighting. Since the current relighting
initiative was intended to save energy, the present research was designed to assess its effect on
the lighting quality in the building.
When lighting systems are retrofitted to conserve energy, it is critical to at least maintain, if not
improve, the quality of the lighting. One of the most effective means for determining how the
quality of the lighting changes as the result of a relighting effort is to ask the opinions of
occupants of the lighted environment. The objective of the present research was to determine
whether the relighting improved the overall lighting quality in the building and the occupant
attitudes toward the lighting. A further objective was to measure the physical lighting
characteristics before and after the retrofit. The overall energy consumption of the lighting
system was evaluated by others, and so was outside the scope of this effort.
Post-occupancy evaluation techniques, including questionnaires and physical measurements, were

used to evaluate the effectiveness of the relighting. These two techniques allow identification
of relationships between subjective measurements of the environment provided by the occupants
and objective physical measurements of the same spaces. Combined, they allow a determination
of whether the changes made during the relighting improved both the physical conditions and
the occupant response. Environmental conditions of concern in this study included lighting,
temperature, general appearance, and presence/absence of windows. The occupant questionnaire
addressed these areas, although it concentrated on response to the lighting. The physical
measures concentrated on lighting - both task illuminance and room luminance - although
supplementary measures were taken of VDT’s and thermal conditions.
3
A representative sample of 350 offices was selected for evaluation by DOE officials after
consultation with NIST staff. This sample proportionately reflected the types of offices in the
whole building based on their location within the building and presence of windows. The final
sample included windowed and windowless offices (to assess any daylight contribution) from the
basement to eighth floor, with windows facing North, South, East, and West. The sample
included conventional as well as open-plan offices.
The two-part post-occupancy analysis was executed twice to collect the necessary data. The
protocols for the questionnaire survey and the physical measures of the occupied spaces are
described in the following sections (and presented in Appendix A). This analysis was first
implemented in March 1993, before the relighting was initiated, and again in December 1993,
two months after the relighting was completed.
2.2.1 Questionnaire Survey
The questionnaire survey was developed based on that used by Collins, et al, (1989) and Rubin
and Collins (1988) for their evaluation of U.S. Army field stations but modified to focus in
greater detail on the lighting. The three-part questionnaire generally covered respondent
attitudes toward lighting, air quality, temperature, VDT’s, space, noise, windows, facility
appearance, and job satisfaction, as well as respondent’s length of time at the facility and general
demographic information. (The questionnaire is provided in Appendix A.) The first two parts
of the questionnaire consisted of statements printed on a standard machine-readable form. The
subjects responded by blackening the circles which represented their responses. In Part A,
participants indicated the degree to which they agreed with the statements, and in Part B they
rated the quality of particular environmental attributes. Part C allowed the subjects to provide
additional written information which could not be easily collected using the machine-readable
forms.
The questionnaire was distributed to the predetermined office spaces and included a business
reply envelope in which the participants could return their completed forms. Although it was
desirable to have paired data; that is, data from the same person before and after the relighting,
this was not possible for two reasons. First, employees moved from one office to another during
the course of the experiment and second, it was critical to maintain the anonymity of the
participants. In spite of these restrictions, there was a concerted effort to ensure that the same
offices were included in both questionnaire surveys. In this way it was possible to solicit
responses from at least some of the same people. In fact, 30 percent of the participants in the
post-retrofit survey volunteered that they had completed the same survey prior to the retrofit.
Physical measurement data were collected from virtually the same offices.
2.2.2 Physical Measurements
Physical measurements were collected in March of 1993 (after the first questionnaire) and again
in January, 1994 (after the second questionnaire). The physical data collection included a
4
general space inventory as well as measures of lighting (both illuminance and luminance),
temperature, and color. From these data, task contrast was calculated. The protocol used to
collect these data is presented in Appendix B.
Prior to the retrofit, physical measures were made with hand-held instruments. The illuminance
was measured at a number of locations within the space, using a Minolta1 photometer with a
cosine-corrected diffuser and a photopic response filter. Illuminances were recorded for six
locations on the main worksurface corresponding to a 3x2 grid. Vertical surface (wall)
illuminances were measured at ten locations within each work space to characterize two
horizontal bands on vertical surfaces within the space. The first band of five illuminances was
measured approximately one meter above the floor and the second band was measured at a
height of about 2 meters above the floor. Within each band five illuminances were measured,
one to the immediate left and right of the observer seated at the desk, one straight ahead, and
one approximately 45 degrees to the left and right of straight ahead. In addition, several
illuminance measures were recorded to characterize the VDT work station. At the work station,
the illuminance of the source document was recorded, as well as the illuminance of the keyboard
and the VDT screen itself. The source document location was assumed to be to the left or right
of the VDT and in a vertical or horizontal position as dictated by the presence of a document
holder. The final illuminance recorded for each office was for a distinct position on the wall
and was later used with luminance measures to calculate the reflectance of the paint used in the
offices.
Luminances were measured using a Minolta Chromameter1 with a field of view subtending a
visual angle of 1° and calibrated to measure photopic luminances. Room surface luminances
were recorded for a pattern similar to the previously described bands of illuminance
measurements and from the viewing position of the office occupant (usually the seated position
behind the desk). In addition, task luminance was measured for two "standard" tasks located
both on the desk and at the position of the assumed source document location at the VDT
workstation. One "standard" task was matte; the other was glossy in nature. The matte task
was a white piece of paper with a black inked letterhead. The glossy task was a page from a
magazine and consisted of a large area of white and a somewhat smaller area printed with black
ink. Each task was placed at the near center of the desk and at the appropriate position at the
VDT work station. The luminance of the white and black areas were recorded and later used
to calculate the task contrast according to the following formula:
1 Brand names are given for the purposes of identification and do not constitute an
endorsement by either the National Institute of Standards and Technology or the Department of
Energy.
5
C = (Lb - LtVLb
where:
Lb = luminance of the background, and

Lp = luminance of the target.
Luminances were also recorded for the brightest spot on the ceiling which was, invariably, an
image of the lamp on the prismatic lens of a luminaire, and for the dimmest spot which was
chosen to be at a location between luminaires. If there were windows, the luminances of the
sky and of a nearby building were recorded. Lastly, the luminance of a distinct spot
corresponding to the earlier described illuminance measurement was recorded to allow
calculations of the reflectance of the painted walls.
Other measures recorded in each office included CIE2 x,y color coordinates for a single point
on the wall as measured with the Chromameter; and temperature and relative humidity which
were measured with a Solomat multi-channel modometer (2016) fitted with a platinum
thermohygrometer.
After the retrofit, the same measures were recorded with a few modifications. To ease the data
collection, a portable automated data collection device was used to collect most of the
illuminance and luminance measures, as well as temperature and relative humidity. This
collection device consisted of a 14 channel datalogger attached to a pole mounted on a wheeled
base. Twelve photocells and a temperature and relative humidity probe were attached to the
datalogger. Each of the photocells had a cosine-corrected diffuser and a photopic response filter.
Eight of the photocells measured illuminance and four photocells measured luminance. Three
of the four photocells used to measure luminance were identical to those used to measure
illuminance except that a 2.6 cm tubular extension lined with black felt and having an inside
diameter of about 2.2 cm was added to the front of the photocell to limit the viewing angle of
the photocell to about 50°. Finally, the fourth photocell was calibrated for photopic viewing
conditions and had a field of view of about 15°. It was mounted on a motorized turntable
attached to the top of the pole at a height corresponding to seated eye level.
The twelve photocells were arranged in four different configurations. The first configuration
was a folding "X" frame in which one illuminance probe was attached to each end. This
configuration was used to measure the illuminance at four points on the worksurface. The
second configuration consisted of a meter stick with an illuminance probe attached to either end.
The third configuration was used to collect photometric information at the VDT workstation.
It consisted of a rectangular frame which hung from the top edge of the VDT monitor. Attached
to the frame were two illuminance probes and the three luminance probes fitted with the tubular
CIE stands for Commission Internationale de l’Eclairage or the International

Lighting Commission which defines color in terms of x and y coordinates.
6
extensions. The luminance probes were positioned to measure the luminance to the left, right
and above the monitor. One illuminance probe was positioned to measure the illuminance on
the VDT screen, while the final illuminance probe was used to measure the illuminance on the
keyboard. In the fourth configuration, the final luminance probe was attached to a motorized
horizontal turntable mounted at the top of the pole.
The datalogger was programmed to collect the data automatically in three different stages. In
the first stage, the datalogger replaced the chair behind the desk and the folding "X" was placed
on the worksurface. Upon launching the collection program, the four illuminances and
temperature and relative humidity were recorded. In addition, the rotating luminance probe
mounted to the top of the pole began a 180° scan stopping every 45° to measure the luminance
of the surrounding surfaces. The first measure was recorded for a location immediately to the
left of the position normally assumed by the employee seated at the desk, and subsequent
measures were recorded for positions located in front of them. The second stage of the data
collection involved holding the meter stick with the two illuminance probes in a vertical position
with the lowest photocell one meter above the floor for each of the five positions described
earlier. The third and final phase of the automated data collection involved hanging the
rectangular frame containing the three luminance and two illuminance probes on the VDT
monitor.
The datalogger stored the measurements as uncalibrated voltages in a storage module. At the
end of each day, the data were uploaded onto a personal computer where it was later imported
into a spreadsheet and converted to meaningful data by applying the appropriate calibration
factors. The calibration factors for all of the probes were provided by the manufacturer
assuming that the cells were to be used for illuminance measurements. However, since we
configured the luminance probes in our laboratory, we also calibrated them using a standard
luminance source. The standard source is an integrating sphere with a small opening in the wall
and is internally illuminated by a fiber optic source. The luminance probes were mounted along
the axis perpendicular to the opening in the sphere such that the opening of the tubular extension
was aligned with the opening in the sphere. This ensured that the luminance probe was indeed
exposed to a uniform luminance and that non-uniform areas were not included in the field of
view.
In summary, then, the differences in measurement techniques and measurements before and after
the retrofit can be described as follows: work surface illuminance was characterized by six
measurements before the retrofit and four measurements after. Prior to the retrofit, room
surface luminances were characterized by a series of ten measurements made with a luminance
meter with a very small spot size (1°) while the post-retrofit room surface luminances were
characterized by five wide angle measurements. Otherwise, the measurement procedures were
similar for the two data sets.
Prior to the retrofit, physical measures were taken in a representative 100-office subsample of
the 350 offices selected to receive questionnaire surveys. In this way the physical measures
were recorded for offices from which questionnaire data were likely to be collected. This
7
provided some assurance that the physical measures reflected the conditions actually experienced
by the occupants. After the retrofit, physical measurements were taken in only 75 offices
because virtually all of the renovated work stations throughout the building were identical.
In addition, since the spectral characteristics of the lamps changed from cool white before the
retrofit to triphosphor after the retrofit, a set of color samples was measured before and after
the retrofit to document the resulting color shifts. CIE x, y coordinates were measured with the
Chromameter for a series of Munsell color chips found inside the back cover of the IESNA
Handbook (Kaufman, 1972). The color measurements were recorded in three separate
windowless offices before and after the retrofit.
8
3. Questionnaire Results
3.1 General
A total of 464 questionnaires were completed by Forrestal Building occupants, 244 before and
220 after the lighting system was retrofitted. Of the 220 respondents to the post-retrofit survey,
64, or about 30 percent, also completed the survey prior to the retrofit. All participants3
engaged in typical office tasks including routine use of VDT’s, paper tasks, phone and face-to-
face conversations.
The data for each questionnaire item collected before and after the retrofit were inspected for
anomalies and analyzed using an analysis of variance (ANOVA) procedure with two dimensions:
sequence - before and after the retrofit; and window orientation - north, south, east, west, and
none. Since all levels of the window orientation factor did not have equal n’s, as shown below,
the particular ANOVA procedure chosen for the analysis assumed unequal sample sizes. In
addition, statistical significance was based on an alpha of 0.05. The results of the analysis are
tabulated at the end of this section.
The number of respondents having a particular window orientation is shown below, with the
percentage of respondents shown in parentheses.
Before After
North 29 (12 percent) 27 (12 percent)

South 16 (7 percent) 14 (6 percent)
East 23 (10 percent) 21 (10 percent)
West 16 (7 percent) 18 (8 percent)
None 155 (65 percent) 138 (63 percent)
The statistical analysis of the data allows two important general observations. First, the items
addressing lighting concerns were usually statistically significant when before and after retrofit
conditions were compared (termed "sequence" in this report). Conversely, while non-lighting
concerns were statistically significant for the window orientation factor, they were not for the
sequence factor. This observation is fundamentally important because it suggests that the
participants responded to the changes in the lighting system rather than to other environmental
variables. Secondly, the statistically significant items regarding lighting showed a positive shift
in response as a function of sequence. That is, occupants responded more favorably to the
retrofitted lighting system than to the original system.
The participants’ ages ranged from less than 20 years to more than 60, with a
median age of about 45 years.
9
The figures and tables in section 3 summarize the questionnaire results in terms of the mean
ratings given for selected questions. In the following sections, figures are presented for items
which were statistically significant. The questionnaire items that were statistically significant
for the window orientation factor were further analyzed to determine which means were different
from the others using the Newman-Keules Multiple Comparisons technique for unequal number
of participants and unequal variances (Howell, 1982). Questions for which no statistical
significance was obtained are discussed in the text, and tabulated at the end of this section, but
not presented graphically.
3.2 Subjective Response to Lighting
Figure 1 presents graphs containing the responses to three questions - amount of light to see
clearly, dimness of the light, and overall amount of light - in terms of ratings of the quantity of
light in the office before and after the retrofit. In this and subsequent figures, the black bar
represents the mean ratings before the retrofit, while the gray bar depicts mean ratings after the
retrofit. For these figures, ratings near 3 can be interpreted as neutral. All differences shown
in the figures were statistically significant at or beyond p<.05, according to the statistical
analysis. The numeric data on which these conclusions are based are tabulated at the end of
section 3. Inspection of the three graphs shown in figure 1 demonstrates that attitudes toward
the lighting were significantly more positive following the retrofit. Occupants indicated that
they could see more clearly, that the lighting was less dim, and that they had a greater overall
amount of light. (Occupants disagreed that their lighting was too dim - a double negative which
we interpret as meaning that they found their light to be sufficiently bright.)
Questions about quantity of light for which there was no significant change in the response were
"the lighting in my office is too bright" with mean ratings4 of 3.78 before and 3.67 after,
indicating that occupants somewhat disagreed with this statement; and "I am satisfied with the
lighting in my work space" with mean ratings of 2.85 before and 2.65 after, indicating that
occupants somewhat agreed with this statement. In both cases, occupants were slightly more
positive after the retrofit about the lighting conditions in their offices (although not statistically
significantly).
The difference between these two means is NOT significant for these data and for
all subsequent presentation of means within the text in section 3.
10
Figure 1. Ratings of the Quantity of Light as a Function of Sequence
Strongly
5 Poor
Disgree
Somewhat
4 Fair
Disagree
Neutral 3 Adequate
Somewhat
2 Good
Agree
Strongly
Agree 1 Excellent
The amount The lighting Amount of
of light on my in my office lighting.
work allows is too dim.
me to see it
clearly.
Legend
Before
After
11
Figure 2 presents data related to the appearance of the lighting system in the form of answers
to six questions. The first four asked occupants to agree or disagree with a statement about their
lighting, while the second two asked occupants to rate the appearance of the lighting on a five
point scale. As can be seen in figure 2, occupants found their spaces to be less poorly lit, more
pleasing, less glaring and the pattern of lights less bothersome after the retrofit. They also rated
the appearance of their lighting system as better, along with the quality of the lighting.
In data not presented graphically but contained in table 1, occupants disagreed that the lights in
their workstations made a loud buzzing noise, either before or after the retrofit (4.22 vs 4.46).
Interestingly, they found no differences in the lighting of the conference rooms (2.97 vs 2.93)
or restrooms (2.9 vs 3.06) - spaces that may not have been relamped at the time of the second
questionnaire. They did, however, rate the relamped hallways as having better lighting quality
(3.47 vs 3.10).
Data for the four questions related to VDT’s and presented in table 1 demonstrated no significant
differences before and after retrofit. Thus, occupants did not find that their lighting washed out
the computer display (mean 3.7) or created distracting reflections in the screen (mean 3.3).
They also reported that their screens did not have anti-glare devices (mean 4.1), but that they
could easily adjust screen brightness (2.11 vs. 1.98). They also reported no problem with
shadows on their work because their body blocked the light (3.85) and they indicated that they
could easily adjust their viewing distance to the screen (2.66 vs 2.7).
Figure 3 presents data which show the effect of workstation orientation (north, south, east, and
west or none). These data reinforce the importance of the window in determining occupant
response. As noted earlier, the effect of the lighting retrofit sequence was not significant for
these questions. Figure 3 indicates that occupants largely disagree with the idea that they would
mm off some lights, although those in east-facing offices were more neutral on this question.
All occupants, except those in west-facing offices, somewhat agreed with the idea of having an
adjustable light on their desks. Occupants found no significant difference in the ability to adjust
their lights with non-significantly different mean ratings of 4.35 before and 4.29 after.
In figure 4, data for three questions reflect occupant attitudes toward their space, again as a
function of their window orientation. The last bar of each set contains the responses of those
in windowless offices - all of whom found their spaces more confining and less adequate for
their jobs than those in offices with windows. Occupants in offices with windows also found
the amount of space for their work to be better than did occupants of windowless offices (even
though the amount of space was similar in both types of offices -14.4 m2 vs about 12.6 m2 or
160 ft2 vs about 140 ft2). Occupants found their spaces to be somewhat more spacious after the
retrofit (3.87 vs 3.64), but again this difference was not significant.
12
Figure 2. Ratings of the General Appearance of the Lighting System as a
Function of Sequence.
13
Figure 3. Ratings of Control of Light Sources as a Function of Window
Orientation.
Strongly
Disgree 5
Somewhat
Disagree 4
Neutral 3
Somewhat
Agree 2
Strongly
Agree 1
If I could, I would I would like to have an
turn some lights off adjustable light on my
since they are too bright. desk.
Legend
North
South
West
NMi None
14
1
Figure 4 Ratings of the General Amount of Space as a Function of Window

Orientation.
Strongly
3.3 Thermal Environment
Figure 5 presents data related to the thermal environment. Again, the main effect is that of the
window - rather than the lighting retrofit. Occupants in windowless offices generally found the
air circulation to be less adequate, less humid, and hotter than occupants in other orientations,
although there was a slight tendency for the approximately 15 people in the south-facing offices5
to find the air circulation to be less adequate, and hotter than occupants of other, windowed
offices. Occupants generally rated the ventilation and air circulation to be adequate to fair -
again with those in windowless offices being somewhat less satisfied, closely followed by those
in south-facing offices. Occupants did not find their offices to be too cold, as evidenced by
almost identical mean ratings (3.3) before and after the retrofit, although they did find the room
temperature to be only fair with ratings of 4.0 and 3.93 for the two questionnaires.
3.4 General Appearance of the Offices
Figure 6 presents data from four questions related to the general appearance of the space. Again
the presence of a window showed the strongest effect. In general, occupants were relatively
neutral to slightly negative about the overall appearance of their offices. Those in windowless
offices found the overall appearance of their offices to be less pleasing; were slightly less
satisfied with their work spaces; and found them to be less colorful and less stimulating than
occupants in windowed offices.
For the questions about office appearance, the biggest effect is shown by the "windowless"
conditions, with much less difference between the various window orientations. These data
suggest that the presence of a window played the biggest role in determining occupant response
to the general appearance of their offices. When sequence effects are considered, no significant
differences emerge in attitudes toward the colors in the offices, with occupants rating the
"naturalness" of their colors as 2.7 before and after the lighting retrofit, and the color of
furniture and objects as 3.4 before and after. There was also no significant difference in the
appearance of the space before and after (mean ratings of 3.5). They also believed that they had
enough space to perform their job, again with no significant difference between retrofit
conditions (2.49 vs 2.51). The spaces were rated as being somewhat less than pleasant with no
change between retrofit conditions (3.59 vs 3.55).
South-facing offices had Venetian blinds for occupants to use to reduce direct
solar radiation.
16
Figure 5. Ratings of the Thermal Environment as a Function of Window
Orientation.
17
Figure 6. Ratings of the Overall Appearance of the Office as a Function of
Window Orientation.
West
CD
Strongly
O)
to to
5 £
-g hO
l 8 a> ><
O £o 5 2 £
CL)
18
Figure 7 presents data on the attitudes toward noisiness of the office space. Occupants of south¬
facing offices found them to be somewhat noisy, while those in windowless offices found them
to be less noisy. Occupants generally agreed that they had reasonable conversational privacy
with no significant differences in their ratings for this question (3.63 vs 3.69) between
conditions, and did not feel that noise kept them from doing their job well (3.26 vs 3.14).
Occupants also reported no significant differences in the condition of their desks and chairs (3.0
vs 3.15); maintenance of the space (3.35 vs 3.51); and cleanliness of the space (3.19 vs 3.41).
Three questions about general morale indicated that the participants had very positive attitudes
toward their work, with no significant differences between the relighting conditions or window
conditions. Respondents clearly believed that their work was important to the operations of their
office (1.62 vs 1.56); were satisfied with the quality of their work (1.74 vs 1.78); and generally
disagreed with the idea that the work atmosphere was tense (2.87 vs 2.99). These data are
among the most positive contained here, indicating no adverse effects of the relamping.
Table 1 summarizes the results of the statistical analysis for all the questionnaire data. This
analysis was based on a two-factor Analysis of Variance for each question where the two factors
were sequence (Before vs After) and window orientation (North, South, East, West, and None).
The first part of the table. Part A, summarizes those items that were statistically significant for
the sequence factor - implying that the differences in participant responses between retrofit
conditions were not due to chance. The last column provides a descriptive statement about the
basic trend of the data. The second part of the table, Part B, reports the items that were
significant for window orientation. The means were later analyzed with a Newman-Keules
Multiple Comparisons test based on unequal n’s and unequal variances (Howell, 1982) to test
which means were statistically different from the others. The means that share a common
underline are statistically the same but different from those which do not share the same
underline. The third part of the table, Part C, lists the items that were not statistically different
along any factor. For these items, the Grand Mean ± 1 standard error of the mean is reported.
Part D, the last part of the table, lists those items that exhibit an interaction between the two
factors. In Table 1 each item includes a code in the left-hand column which indicates the type
of response requested of the occupant. This code correlates the verbal response to the reported
numerical ratings which are included at the top of Table 1.
3.5 Reasons for Choices
Occupants were also asked to indicate which four aspects of their offices that they would change
if they could. Figure 8 summarizes the data from before and after the retrofit in a stacked bar-
graph format. In this graph, the choice is presented along the abscissa, while the percentage of
respondents who selected that choice is presented for each of the four possible choices. Figure
8 indicates very clearly that the most desired choices were "more comfortable day-to-day
temperatures"; "improved air circulation"; and "view out/daylight" with relatively little change
between retrofit conditions. The only change between retrofit conditions occurred for "improved
lighting" which was selected much less frequently after the retrofit, demonstrating that the
relighting had achieved one of its goals.
19
Figure 7. Ratings of the Noisiness of the Space as a Function of Window
Orientation.
Poor 5
Fair 4
Legend
North
South
East
West
None
20
Figure 8. The Percentage of Respondents That Chose the Proposed Hypothetical
Changes to Their Office
Choice 3
Other
View out/daylight
More adjustable chair
Adjustable task lighting
More frequent cleaning
More comfortable furnishings
More surface area

for work
Improved air circulation
Less noise
Improved lighting
Change in color of
furnishings
More privacy
More comfortable
day-to-day temperatures
r
o o o © ©
in CO CN o
21
Table 1. Summary of Questionnaire Results for Sequence and Window Orientation
Scale
Item Code 1 2 3 4 5
Strongly Somewhat Somewhat Strongly
A: Agree Agree Neutral Disagree Disagree
B: Excellent Good Adequate Fair Poor
PART A: Items that are statistically different for the Sequence Factor.
Mean Mean Difference P

Code Question Before After in Means value Trend
A The amount of light 2.11 1.73 0.38 0.00 Better

on my work allows me after
to see it clearly.
A The ceiling lights in 3.26 3.54 -0.29 0.02 Less

my work area are glare
glaring after
A The light in my office 3.08 2.73 0.35 0.00 Better

is pleasing after
A The irregular pattern 3.26 3.76 -0.50 0.00 Less

of lighted ceiling lights bothersome
bothers me after
A My space is poorly 3.51 4.03 -0.52 0.00 Better

lighted lighting
after
A The lighting in my 3.76 4.10 -0.34 0.00 Brighter

office is too dim after
22
Amount of lighting 2.90 2.57 0.33 0.00 Better
amounts
after
The quality of the 3.47 3.10 0.37 0.00 Better after

lighting in the corridors
The quality of the light 3.14 2.79 0.35 0.00 Better

in my work space quality
after
PART B: Items that are statistically significant for the Window Orientation Factor:
A I would turn some lights off since they are too bright
East None West South North

3.49 3.90 3.97 4.07 4.32
General disagreement implying not enough light.
A I would like to have an adjustable light on my desk
South North None East West

2.53 2.54 2.62 2.91 3.32
General agreement although East and West exposures less.
A My space is confining
None East North South West

2.65 3.25 3.36 3.53 3.82
General agreement with those having windows agreeing less
23
A The space is adequate for my job
South West North East None

2.41 2.47 2.61 2.66 2.94
Generally agreeable (None - windowless - less agreeable)
B Amount of space for your work
East West North South None

2.84 2.85 2.87 2.88 3.34
Generally Good, with Windowless Adequate to Fair
A The air circulation in my office is adequate
West East North South None

3.15 3.34 3.45 4.00 4.23
General disagreement. South and None disagree more strongly.
A My office is humid
None South North East West

3.20 3.60 3.75 3.86 3.85
Generally disagree though S and None were more agreeable
24
A My office is hot
None South East North West

2.42 2.53 3.02 3.20 3.25
Generally agreeable especially for South and None
B Ventilation and air circulation

3.53 3.82 3.88 4.33 4.48
West, East, North adequate to fair; South and None fair to poor
A The overall appearance of offices is pleasing

3.12 3.27 3.36 3.40 3.74
General disagreement though those with windows were more pleased.
A My office is colorful
West South North East None

3.41 3.47 3.55 3.70 4.03
General disagreement though those with windows were more agreeable
25
A Overall I like my work space and furniture

2.64 2.80 2.87 2.88 3.35
General agreement with exception of those without windows
A My space is stimulating
West North East South None

3.42 3.53 3.55 3.60 4.12
General disagreement though windowless greater disagreement
B Noisiness of the space
East West South North None

2.95 3.18 3.23 3.35 3.51
East good to adequate, others adequate to fair
B How many hours do you work at a VDT per day
North East South West None

2.05 2.40 2.50 2.56 2.69
Those with windows less likely to work at computers
26
B How many hours do you spend reading and writing
West None East North South

2.71 2.77 2.79 3.20 3.20
North, South windows more likely to spend more hours reading

and writing.
PART C: Items that are not statistically significant for any Factor. Grand Means ± 1
SEM is reported.
A I create shadows on my work because of my body

3.85 ± 0.059 Disagree
A The lighting in my office is too bright

3.73 ± 0.06 Disagree
A I am satisfied with the lighting in my work space

2.74 ± 0.06 Neutral to Agree
B Ability to adjust light for work

4.33 ± 0.05 Fair to Poor
A The amount of light in my office washes out the VDT

A Ceiling lights create distracting reflections in my VDT

3.38 ± 0.06 Neutral to Disagree
A My computer has an antiglare device

A I can easily adjust the brightness of my VDT

2.04 ± 0.05 Agree
A The colors in my office appear natural

2.78 ± 0.056 Agree to Neutral
B Color of furniture and objects

3.44 ± 0.05 Adequate to Fair
27
A My space appears spacious
B The appearance of the workspaces and offices

B Maintenance of the space

B Condition of desks and chairs

3.08 ± 0.05 Neutral
A My work is important to the functioning of this office

1.59 ± 0.43 Agree
A I am satisfied with the quality of my work

1.77 ± 0.04 Agree
A The work atmosphere in my office is tense

2.93 ± 0.06 Neutral
A My office is cold
3.32 ±0.06 Neutral to Disagree
A The lights in my work area make a loud humming sound

B The quality of lighting in conference rooms

2.96 ± 0.05 Neutral
B The quality of lighting in the restrooms

3.02 ± 0.05 Neutral
B Conversational privacy
28
B Cleanliness of the space
PART D: Items that show an interaction between the two factors:
A I have enough space to perform my job INTERACTION

A Noise keeps me from doing my job well INTERACTION
A I can easily adjust my viewing distance to the VDT INTERACTION
B Pleasantness of the space INTERACTION
B Room temperature INTERACTION
29
4. Measurement Results
4.1 Lighting Measurements
Using a procedure similar to that given in Appendix B, a total of 100 work stations were
examined before and 75 after the relighting. All the physical measurement data recorded for each
office were entered into a spreadsheet for easy manipulation. A number of summary measures,
such as task contrast, mean worksurface illuminance, and mean wall luminance, were created.
These data were analyzed similarly to those presented for the questionnaire results, using a two-
factor analysis of variance assuming unequal sample sizes and variance. As in that analysis, the
two factors were sequence and window orientation. The results are presented in Table 2.
As with the questionnaire analysis, the following figures present the measurement data that were
significant for the sequence or window factors. Figure 9 presents the distribution of illuminance
at the primary work station for five measurement points located at two distances (lm and 2m)
above the floor before and after the relighting. The upper portion of figure 9, 9a, presents
illuminance data for all offices; while the lower portion, 9b, presents data only for windowless
offices. Inspection of the open diamonds in figure 9a indicates clearly that illuminance was
greater after the relighting than before. The increase was greatest for measurement points located
2m above the floor - which went from about 290 lux to about 450 lux after the relighting. A
similar change can be seen in figure 9b, where the illuminance for points located 2m above the
floor increased significantly after the relighting. These shifts indicate that there was little
luminance gradient within the offices before the relamping - meaning that the upper portion of
an office had relatively less luminance than desirable. Figure 9b also provides an indication of
the role of daylight in the overall illuminance in the offices since the 2m measures increased only
to 400 lux for windowless offices after the retrofit as compared with 450 lux for all offices.
Figure 10a presents data for mean horizontal illuminance on the workplane before and after the
retrofit. Before data are shown in black and after in gray. Figure 10a further separates the data
into categories according to window orientation. The central line at 500 lux was the target
minimum value desired by DOE for the relighting. This figure indicates clearly that east-facing
and windowless offices had levels below the target minimum maintained horizontal illuminance
before the relighting. All offices had levels above this target minimum following the relighting,
although the level was lowest for offices with no windows. In addition, after the retrofit, offices
with windows had mean horizontal illuminance at or above 650 lux; that for the windowless
offices was about 550 lux. Figure 10b presents mean vertical illuminance on the wall for all
offices, before and after the retrofit. This figure summarizes the data for all windowed conditions
and orientations. It confirms that wall illuminance increased significantly following the retrofit,
with an increase from about 260 lux to about 340 lux - for all offices.
30
Table 2. Summary of Physical Measurement Data
PART A: Items that were statistically significant for the Sequence Factor.
Mean Mean P
Measurement Before After value
Mean standard deviation of wall 74 108 0.01

illuminances in each office
Mean CIE y-coordinate for office walls 0.425 0.413 0.00
Mean luminance of darkest spot 40.6 47.1 0.03

on ceiling (cd/m2)
Mean luminance of brightest spot 3094 4712 0.01

on ceiling (cd/m2)
Average wall illuminance (lux) 269 339 0.04
Number of lamps per square meter 0.69 0.42 0.01
Relative Humidity (percent) 28 14 0.00
Mean file length 0.94 10.1 0.00
PART B: Items that were statistically significant for the Window Orientation Factor:
Source document illuminance (lux)
None East South West North

385 523 531 603 691
31
Luminance ratio of brightest spot to dimmest spot on ceiling (cd/m2)
North West South East None

62 85 88 92 104
Mean work surface illuminance (lux)
None East South West North

491 537 602 654 719
PART C: Items that are not statistically significant for any Factor. Grand Means ± 1
SEM are reported.
Mean background luminance of matte paper task (white)

196.5 ± 46.5 (cd/m2)
Mean task luminance of matte paper task (black)

22.4 ± 3.3 (cd/m2)
Mean task contrast of matte paper task

0.87 ± 0.01
Mean background luminance of glossy paper task (white)

184.7 ± 47.5 (cd/m2)
Mean task luminance of glossy paper task (black)

19.0 ± 6.6 (cd/m2)
Mean task contrast of glossy paper task

0.91 ± 0.01
Mean CIE x-coordinate for office walls (color)

0.418 ± 0.001
Mean office area

13.3 ± 0.37 m2
32
Mean desk area
1.4 ± 0.02 m2
Mean reflectance of office walls (rho)

0.77 ± 0.004
PART D: Summary Descriptive Statistics
Before After
Percentage of offices with computers 97 97
Percentage of computers with adjustable screens 96 99
Source document location

Left of computer horizontal 31 42
Left of computer vertical 10 3
Right of computer vertical 7 6
Right of computer horizontal 52 49
Percentage of "Open Plan" Offices 10 8
Percentage of offices having stated number of occupants

One 87 91
Two 11 8
Three 1 1
Four 1 0
Percentage of offices with pictures or posters 87 95
Percentage of offices with personal effects on the desk 77 93
Percentage of offices with personal fans 35 37
Percentage of offices with space heaters 3 3
33
Figure 9 Distribution of Measured Wall Illuminances as a Function of the
Illuminance (lux)
Presence of a Window in the Office.
1 2 3 4 5
Measurement Points
Desk
9a. Illuminance Distribution in All Offices
Measurement Points
Legend
-A-Before
Illuminance (lux)
-&— After
Measurement Points
9b. Illuminance Distribution in Windowless Offices
34
Figure 10. Mean Horizontal Illuminance as a Function of Window Orientation
and Sequence and Mean Vertical Illuminance on the Wall as Function
of Sequence.
10 a. Mean Horizontal
Illuminance on the
Workplane
10 b. Mean Vertical Illuminance on the

Wall for all Offices Before and After
the Retrofit.
35
Figure 11 describes the pattern of lamps and luminances in the offices before and after the
relighting. Figure 11a, which presents the mean density of lamps, indicates that there were
fewer lamps per square meter after the relighting than before. Figure lib amplifies these data
by indicating that, although there were fewer lamps energized, these lamps were located in a
far greater percentage of luminaires. Before the relighting only about 72 percent of the
luminaires were energized; after, almost all (about 98 percent) the luminaires were energized.
Before the relighting, many of the two-lamp luminaires had been delamped. After, each
luminaire had one energized lamp. This change to a greater number of energized luminaires
meant that the "patchwork" appearance of the ceiling - with a few very bright luminaires
contrasting with some very dark (unlit) areas - was lessened. Figure 11c indicates that the
mean luminance of the brightest spot on the ceiling increased significantly - from about 3000
cd/m2 to about 4600 cd/m2. This was likely due to the higher lumen output of the new lamp
compared to the old and the fact that the brightest spot on the ceiling was the image of the
lamp through the prismatic lens of the luminaire. In addition, fixtures were cleaned during
the relighting which would also have increased their overall brightness. Finally, figure lid
presents the mean ratio of highest ceiling luminance to lowest, which also increased after the
relighting, again reflecting that the lamp images were brighter. Consideration of the four
graphs in figure 11 suggests that the pattern of luminances in the office had become more
regular and that the upper walls and ceiling had become generally brighter after the relighting.
Figure 12 presents CIE x,y coordinates for color samples measured before and after the retrofit.
The set of color samples, used to approximate interior accent colors, were Munsell color chips
from the IESNA Handbook (1972). Figure 12 indicates that the gamut of colors (Boyce, 1977)
had increased after the retrofit, particularly in the red-purple range. The change from cool
white to T-8 triphosphor lamps (with higher color rendering and color temperature) would
account for this increased gamut of colors.
36
Figure 11. Pattern of Lights and Luminance as Function of Sequence
37
Figure 12. Shifts in CIE x,y Coordinates for Color Samples as a Function of
Sequence.
0.500 q .Green Yellow

Green .
O*
> Yellow
0.450 :
Blue-Green
b* Yellow-Red
Color Name Munsell Value
0.400 - $N'eutral Yellow 5Y 8.5/10
Yellow Red 5YR7/10
\ Red
Red 5R6/10
Blue^ Red Purple 5RP6/10
y 0.350 -
\
Purple
Red-Purple Purple
Neutral
Green Yellow
5P6/8
N6
5GY 8/8
Purple-Blue Green 2.5G 7/8
0.300 : Blue Green 5BG 7/6
Blue 7.5B 6/8
Purple Blue 5PB 6/10
0.250 -
0.200 H—i—i—i—i—r—r—i—i—i—i—i—i—i—i—i—r
Legend
0.200 0.300 0.400 0.500 0.600 A Before
O After
x
38
4.2 Other Environmental Measures
Figure 13 summarizes the temperature and humidity data. It is important to realize that the
significant shifts observed here are most likely due to changes in the external weather: the data
in December 1993/January 1994 (After) were collected during a severe cold snap on a series
of overcast, snowy days. The data in March of 1993 (Before) were collected under sunny
conditions when the sun angle was low enough to allow considerable solar heat gain. Figure
13a reinforces this point with a significant decline in temperatures observed in all offices
(regardless of window orientation) after the retrofit. The decline was greatest for south-facing
offices, suggesting that solar heat gain was a key cause of the March 1993 temperatures. It
should also be noted that the mean temperatures observed for the south-facing and windowless
offices were slightly above the ASHRAE norm of 23.6° C for winter temperatures. (These
data agree with the fact that occupants tended to rate their offices as warm rather than cold.)
All mean temperatures were well above the ASHRAE recommended winter low of 20° C.
In addition, the offices tended to be rather dry, particularly in the winter. Thus, figure 13b
demonstrates quite a drop in relative humidity between the two measurements, with winter
levels near 14 percent and spring levels near 27 percent. Again, the winter humidity levels are
below those recommended by ASHRAE for comfort. Finally, figure 13c plots the percentage
of offices having fans as a function of window orientation. Two locations emerge as having
a significantly greater percentage of fans (more than 40 percent) - south-facing and windowless
offices. This finding reinforces the idea that these offices were hotter and less comfortable
than offices in other locations. Since fans are not provided by building management so that
employees must supply their own, these data further support the idea that these offices were
warm enough to force employees into action. An additional contributing factor to the
decrease in temperatures following the retrofit may well have been the change to more efficient
(cooler) lamps and ballasts which would contribute less heat to the office.
As noted in the text, Figures 9-13 demonstrate significant changes in task illuminance, wall
illuminance, lamp density ambient temperature and humidity following the retrofit. The
changes in illuminance and luminance were positive, and in the direction desired by DOE
management and the retrofit team.
39
Figure 13 Measured Temperatures and Humidity
Temperature (degrees Celsius)
40
5. Conclusions
The data presented in the present report support the hypothesis that the relighting significantly
improved the lighting conditions in the Forrestal Buildings. Consideration of the
questionnaire data indicates that the items addressing lighting concerns were usually statistically
significant - and more positive - when before and after retrofit conditions were compared. On
the other hand, attitudes toward non-lighting components of the building environment (which
were not actually changed as part of the retrofit) did not change significantly between the two
surveys. At the same time, occupants continued to be somewhat negative about thermal and
air quality conditions - thus reinforcing the hypothesis that lighting conditions had truly been
improved. These data indicate clearly that occupants responded to changes in the lighting,
rather than to other environmental variables.
The significant and positive response to the lighting questions is fundamentally important,
indicating that occupants found the retrofit improved the general lighting conditions in their
offices. After the retrofit, occupants found their spaces to be more pleasing, less poorly lit,
and less glaring. They also rated the appearance of their lighting system and the quality of
their lighting as better. Comparison with the earlier data reported by Collins, et al. (1989)
suggests a more positive response to the lighting retrofit in the Forrestal Building. Mean
ratings of lighting satisfaction were higher in the present study than in the previous study,
reinforcing the idea that the lighting conditions had truly been improved. It is of course risky
to compare mean ratings from two different studies, but the trend is toward more positive
response for the Forrestal Building than for the small Army facilities.
Inspection of the physical data shows that the retrofit successfully improved the pattern of
luminance in the offices, with fewer dark spots on the ceiling, and brighter surfaces on the
upper portion of the walls. This is consistent with observations by Loe and his colleagues
(1991) in England who determined that the central 40° field is critical in determining
satisfaction with lighting, and with Collins, et al. (1989) who found that occupants preferred
more balanced luminance distributions in their offices. In addition, the measured task
illuminance increased significantly, to levels above those specified by DOE. It should be noted
that measured illuminance levels were significantly higher for those offices with windows,
indicating that daylight made a significant contribution to the overall illuminance levels and
luminance patterns in the offices. At the same time, occupants were quite negative about the
absence of windows in windowless office areas, making many negative responses in the
comment section of the questionnaire. Their comments reinforced the conclusions drawn by
Collins (1974) and Ulrich (1984) about the importance of windows to people in offices and
other relatively confined spaces.
In conclusion, the retrofit improved attitudes toward the lighting for a significant number of
building occupants. It also improved the lighting for physical measures such as task
illuminance and surface luminance. Thus the goals of the retrofit - to improve the existing
lighting appearance and quality while providing more efficient lighting systems - were met.
41
6. Recommendations for Completing Further Post Occupancy Evaluations
Post-occupancy evaluations (POEs) are very complex projects. They attempt to provide some
understanding of how built environments influence the attitudes and beliefs that occupants
hold about the environment in which they work and how that environment may affect their
job performance. The evaluation could address any single facet of a built environment such
as acoustics, thermal comfort, or lighting; or it could address all of them at the same time.
The present POE has primarily addressed the lighted environment.
Since POEs are complex and address many interacting variables, their success relies
overwhelmingly on the planning and execution of the project, which itself revolves around the
sensitivity of the planned statistical analyses and the potential variability of the data.
The first goal of any project is to accumulate a set of data which will yield statistically
significant results. This does not mean manipulating data to yield desired results, but rather
exercising proper experimental design techniques to yield results that accurately represent
reality. Designing an experiment to yield such a data set is not a trivial task and requires great
skill. However, a few basic principles can serve as a solid foundation. The remainder of this
section describes the fundamental requirements and approaches necessary for executing a POE.
The descriptions are largely based on the experimental instruments used for this analysis
because they have been developed over the course of many projects and have been found
successfully to discriminate occupant responses to changes in lighting.
6.1 Determining the Sample Size and Composition
A successful POE is based on two data sets: one which is a compilation of the responses to
a questionnaire administered to the building occupants, and a second one which contains an
inventory of physical measures of the space. It is usually not cost effective to administer a
questionnaire to all occupants because the building is too large. In this case, a sample must
be selected which is large enough to accurately represent the population of the facility and
permit statistical robustness. At the same time, the sample should be manageable in terms of
minimizing disruptions to the building staff and the difficulty of data collection and subsequent
transcription and analysis.
Sample size is dictated by the anticipated variability of the data to be collected and the number
of factors of interest in the evaluation. But, since the data are yet to be collected and the
variability is therefore unknown, the exact sample size remains elusive. There are, however,
some rules of thumb. First, collect as much data as can be afforded. Excess data are never
wasted but too little data are a complete waste. Generally, small samples provide data which
are highly variable, irregular, and not representative of the population. As such, the statistical
procedures and tests are unable to discriminate what could potentially be very small, but
significant, differences. Larger samples provide more stable data because they include a much
42
larger portion of the population. Therefore, the statistical analyses are less likely to provide
false interpretations of the data.
Secondly, the questionnaire data should be provided by a sample of employees that represents
the work force. This means that the sample should include proportionate numbers of
employees from different administrative levels, such as managers, administrative staff, the main
work force, secretarial staff, and perhaps maintenance staff. This is important because each job
classification has different responsibilities which dictate the exact nature of the visual (or other)
tasks that are completed within a space. The exclusion of a particular group may skew the
data set and provide unwarranted importance to a particular measure which is ultimately
unimportant. Other demographic factors that may be of interest are age distribution and
visual capacities. It would be unwise to collect data about the quantity of light only from
young employees if a large portion of the population is older, since older people may require
more light to complete their tasks. The fundamental point is that the various factors that
affect visual or task performance should be accurately represented in the sample. These factors
include task type, which is determined by job description, the age of the workers, and the
gender of the workers since gender may not be balanced within all job categories.
Just as a number of factors influence the subjective data set, an equally large number of factors
affect the data set for the physical inventory. These include the presence and directional
orientation of windows, the types of electric lighting systems, the geometrical configuration
of the office (whether it is a private office, shared office, or an open office plan), and the
location of the space within the facility. Also, in a multi-use facility where not all the spaces
are conventional offices, the space usage may become a factor of interest. In short, the sample
of rooms should proportionately reflect the nature of the building (or at least areas of interest)
and should be selected from all areas of the facility.
Choosing the samples as representative of the building population or of the facility itself is a
necessary, but not a sufficient, condition for a successful POE. The sufficient condition is
interfacing the requirements for the subjective sample with the requirements based on the
sample required for the physical spaces of interest. After accounting for all the factors of
interest, the next critical requirement for selecting a sample is that it must be a random
selection from the target population. Failure to select a random sample will impose unwanted
constraints on the data and will prohibit a legitimate statistical analysis.
Since the sample size is dependent upon the different factors of interest, it is difficult to suggest
an actual sample size. However, as the number of factors increases, so does the sample size.
Ultimately, the sample size for each data set will likely be at least a hundred. We used 350
in the present study, for example, while Collins, et al. (1989) used about 700 for the
questionnaire sample in their study. For questionnaire data, it is important to inflate the
sample size by about fifty percent over the minimum number of data points needed for a
proper analysis because only about one third of the distributed surveys can be expected to be
returned. The final number of spaces targeted for the physical inventory may be much less
43
than the questionnaire sample if, as in the present study, many of the physical spaces are
virtually identical. We studied 100 spaces initially and 75 after the retrofit, and found this to
be a reasonable number with relative low variability among spaces.
Two further goals in selecting samples which enhance the validity of a POE are to choose the
same physical spaces occupied by survey respondents and to distribute the questionnaire to the
same sample of respondents both before and after the modifications. By measuring the
physical spaces actually occupied by survey respondents, the two data sets will more accurately
reflect the relationship between the occupants and their work spaces. Distributing the
questionnaire to the same sample both before and after the modifications will make it easier
to establish a reliable trend in the data. If at all possible, it is extremely desirable to assign
subjects identification codes which maintain their anonymity and yet permit the data they
supply before and after environmental modifications to be paired. Paired data allows for a
more reliable statistical analysis, especially if the sample size is small. The importance and
significance of this suggestion cannot be overemphasized. It is often very difficult to
implement in practice, however.
6.2 Suggested Questionnaire Contents
The design of a questionnaire is equally important as choosing a good sample. Questionnaires

should not reflect any experimental bias and should contain self-validation structures. In other
words, the questionnaire should address a given number of points. These points should be
addressed by posing a few questions in very generic, non-leading terms. In addition, several
points should be addressed several times in different terminology. In this way, if all the
questions pertaining to a given point reflect similar responses, it can be concluded that the
responses are valid and respond to the theme of the question, rather than being influenced by
the wording of the question. Another important consideration in designing a questionnaire
is that the questions should not "lead" the respondent. Rather, the items of interest should
be disguised by adding a few items of related interest. This deters subjects from fully
understanding the intent of the questionnaire so that they are less likely to provide the data
they think the experimenter wants. This is a critical consideration because, although human
nature is to want to perform well, questionnaires should not be performance measures. Along
the same lines, the precise nature of the experiment should not be described to the subjects
because, again, if they are aware of the goals, they will tend to provide, often unconsciously,
skewed data which is not representative of the conditions. For this reason, physical measures
should be made ideally after the questionnaires have been returned; and the questionnaires
should not be administered immediately after any modifications to the facility.
The questionnaire used in the current POE is included in Appendix A and can be considered
a good model of the types of questions that should be included in a evaluation of lighted
environments. This questionnaire is based heavily on those used in previous post-occupancy
evaluation research conducted by NIST and has been refined with each implementation. The
44
current form is effective in eliciting responses to physical conditions in buildings and is
recommended for subsequent evaluations.
6.3 Suggested Physical Measurements
The physical measurement form suggested for the model POE has been modified and
shortened from that used in the Forrestal Building to reflect more useful measurements. The
original form proved to be overly complex for the type of results we obtained. The revised
form given in Appendix B concentrates on measures of lighting that can be easily taken,
analyzed, and related back to the subjective data. It is similar to forms used in previous POE
research at NIST.
The lighting measures of interest are task illuminance and luminance because these reflect the
quantity of light falling on the task and the task contrast. Both quantities are critical to
assessing task visibility. Wall illuminance and luminance are also important because they very
simply describe the distribution of light within the space, which influences the perception of
the space. The color measurements allow some quantification of how "normal" a space
appears. Depending on the color rendering capability of the lamps chosen, colors may not
appear normal and may, in fact, affect worker productivity if color discrimination is a part of
the task requirements. In the current study, the color measurements quantified the increased
capability of the new lamps to render colors more realistically and also validated occupant
reports of a difference in the appearance of colors in their spaces. It is important to take
physical measures using the same procedure and location in each office. Using standard tasks
and calibrated equipment is essential to obtaining reliable, accurate data. Fundamentally, the
physical measures should, in some way, reflect the items on the questionnaire so that they can
be used to support or refute subjective responses. If they support the responses, it is likely
that the subjective responses are real and are legitimate. If the physical measures do not
corroborate the subjective reports, this is a clear sign that other underlying factors are
influencing the data. In essence, both sets of data validate each other.
6.4 Collecting and Analyzing the Data
After the experimental approach has been developed, administering the questionnaire and
collecting the physical inventory is relatively straightforward. As a basic precaution, any
discussion with the subjects in the building should be kept vague until all of the data is
collected. Any conversations or written material that contain too much information about the
intent of the study and how the questionnaire and physical measures relate to each other could
influence the subjective data. Participants should be debriefed at the conclusion of the study
and given an opportunity to provide further comments, however.
Both sets of data should be collected before and after the modifications to the building are
made. The preliminary, or baseline, data should be collected before the employees are too
involved in the upcoming changes to the space because their knowledge about the proposed
45
changes could influence their response. The post-modification data should be collected a few
months after the modifications are completed. This allows the novelty of the new
environment to wear off and will allow more representative data to be provided.
After the subjective data are collected, they should be analyzed for at least one dimension:
namely, sequence (before and after modifications). In addition, it is helpful to analyze the data
for other suspected significant factors, such as window orientation or office type if the offices
are of mixed types. The statistical analysis should be an analysis of variance because it is a
robust statistical procedure that reduces the possibility that non-significant data are mistaken
for significant data. Finally, multiple comparisons of the significant results should be made
to identify precisely which levels of a particular factor contribute to the significance. Standard
statistical text books, such as Howell (1982), describe these methods. As with the
questionnaire data, the physical measurements should be subjected to an analysis of variance.
Follow-up statistical tests should be used to isolate which levels of a statistically significant
factor are different from one another.
The statistical analyses should be performed using a computer, since hand calculations are
extremely time-consuming. Since statistical analysis packages can be expensive to purchase and
require some expertise to use successfully, it may be advantageous to consult a statistician to
help design the experiment and analyze the data. Statisticians possess the tools and the
necessary knowledge to complete such an analysis. Successful experimental design and data
analysis critically depend on a complete understanding of the assumptions of the statistical tests
to be performed and the proper collection of both physical and subjective data.
Ultimately, the data from both analyses should be compared to determine the overall impact
of any relighting or environmental improvement. In addition, it is helpful to compare the
current findings to the findings from other similar evaluations. This comparison offers the
opportunity to validate the completed study and identify potential problems or areas of the
questionnaire or physical measurement protocols that could benefit from revision.
As stated earlier, POEs are extremely complex studies. They attempt to corroborate changes
in subjective evaluations of built environments before and after an environmental change with
changes in physical measures of the same environment. Because of these complexities, section
6 of this document has attempted to provide some guidance for conducting a POE. While it
does not attempt to outline all of the statistical assumptions and precautionary measures which
are important for a successful POE, it does highlight the basic requirements and provides a
good foundation for completing a POE of lighted environments.
46
7. References
American Society for Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE).

Atlanta, GA. 55-1981.
Boyce, P.R. Investigations of the subjective balance between illuminance and lamp color
properties. Lighting Research and Technology, 1977, 9, Pp. 11-24.
Collins, B.L. Windows and People. Washington, D.C. NBS BSS 70, June, 1975.
Collms, B.L. Evaluation of the Role of Luminance Distributions in Occupant Response to

Lighting, Proceedings. National Lighting Conference, Cambridge, April 1990, Pp. 1-10.
Collins, B.L, Gillette, G.L., Dahir, M.S. , Goodin, P.J. Post-Occupancv Evaluation of Several
U.S. Government Buildings. Gaithersburg, MD: NIST, NISTIR 89-4175, September 1989.
Collins, B.L., Fisher, W.S, Gillette, G.L., Marans, R.W. Evaluating Office Lighting
Environments: Second Level Analysis. Gaithersburg, MD: NIST, NISTIR 89-4069, Apr. 1989.
Collins, B.L. and Rubin. A.I. Analysis of Work Environment Data from Three Army Field
Stations. Gaithersburg, MD: NIST, NISTIR 88-3871, October 1988.
Dillon, R., and Vischer, J.C. Derivation of the Tenant Questionnaire Survey Assessment
Method: Office Building Occupant Survey Data Analysis. Ottawa, Canada: Public Works
Canada, AES/SAG 1-4:87-9, November 1987 (b).
Dillon, R., and Vischer, J.C. User Manual: Tenant Questionnaire Survey. Ottawa, Canada:
Public Works Canada, AES/SAG 1-4:87-8, November 1987 (a).
Flynn, J.E., Spencer, T.J., Martyniuk, O., and Hendrick. C. Interim Study of Procedures for
Investigating the Effect of Light on Impression and Behavior. Tournal of the Illuminating
Engineering Society, 1973, Pp. 87-94.
Gillette, G., and Brown, M. Occupant Evaluation of Commercial Office Lighting: Volume
1., Methodology and Bibliography. ORNL/TM-10264/VI, Oak Ridge National Laboratory,
November, 1986.
Gillette, G. Evaluating Office Lighting Environments: Reference Lighting Power Density Data.
Gaithersburg, MD: NBS, NBSIR 88-3691, January, 1988.
Gillette, G., and Brown, M. Occupant Evaluation of Commercial Office Lighting: Volume
111., Data Archive and Database Management System. ORNL/TM-10264/V2, Oak Ridge
National Laboratory, August, 1987.
47
Hawkes, R.J., Loe, D.L., and Rowlands, E. A note towards the understanding of lighting
quality. Journal of the Illuminating Engineering Society, 1979, Jan, Pp. 111-120.
Howell, D.C. Statistical Methods for Psychology, 1982, Boston:PWS-Kent.
Kaufman, J. E. (Ed). IES Lighting Handbook, Fifth Ed. New York: Illuminating Engineering
Society, 1972.
IES Lighting Handbook: 1987 Application Volume and IES Lighting Handbook: 1984
Reference Volume.. Illuminating Engineering Society of North America, 345 East 47th Street,
NY, NY 10017, 1987, 1984.
Loe, D.L., Mansfield, K.P., and Rowlands, E. Light patterns and their relevance to spatial
appearance and the quality of the lit environment. CIE Proceedings - 22nd Session,
Melbourne, Australia, 1991, Pp. 41-44.
Louis Harris Associates for Steelcase. The Office Environment Index: 1987 Full Report, and
1987 Summary Report. Steelcase, 1987.
Marans, R. W. Evaluating office lighting environments: A further report. Lighting Design and
Application. 1987, 17, p. 32-36, 51.
Marans, R. W., and Brown, M. Occupant Evaluation of Commercial Office Lighting: Volume
II., Preliminary Data Analysis. ORNL/TM-10264/V2, Oak Ridge National Laboratory,
November, 1987.
Meyer, B. Indoor Air Quality. Reading, MA: Addison-Wesley Pub. Co., 1983.
Rea, M. Towards a model of visual performance: Foundations and data. Journal of the
Illuminating Engineering Society. 1986, 15, pp. 41-57.
Rubin, A.I. and Collins, B.L. Interim Survey of Selected Military Building Environments: A
Research Approach. Gaithersburg, MD: NBS, NBSIR 87-3606, August 1987.
Rubin, A.I. and Collins, B.L. Evaluation of the working environment at selected U.S. Army
field stations: Suggestions for Improvement. Gaithersburg, MD: NBS, NBSIR 88-3827, August
1988.
Ulrich, R.S. View through a window may influence recovery from surgery. Science. 1984, 224,
pp. 420-421.
48
Appendix A. Suggested Questionnaire Items
Today’s Date:_
Office Suite: Building:North

South West
Corridor:_
Room No.:_
Part A. Instructions: Please indicate the extent to which you agree with each of the
following statements by circling the appropriate number. If a statement does not
apply to you, do not circle any of the numbers.
Coding Scheme:
1 Strongly Agree
2 Somewhat Agree
3 Neutral
4 Somewhat Disagree
5 Strongly Disagree
1. I have enough work space to perform my job. 1 2 3 4 5
2. The amount of light on my work allows me 1 2 3 4 5

to see it clearly.
3. I create shadows on my work because my 1 2 3 4 5

body blocks the light.
4. The ceiling lights in my work area are glaring. 1 2 3 4 5
5. The lighting in my office is pleasing. 2 3 4 5
6. The colors in my office appear natural. 1 2 3 4 5
7. My work is important to the every day 2 3 4 5

functioning of this office.
49
8. If I could, I would turn some lights in my office 1 2 3 4 5
off because they are too bright.
9. My space is confining. 1 2 3 4 5
10. Noise keeps me from doing my job well. 1 2 3 4 5
11. The air circulation in my office is adequate. 1 2 3 4 5
12. The overall appearance of offices and workspaces 1 2 3 4 5

is pleasing.
13. My office is colorful. 1 2 3 4 5
14. The irregular pattern of lighted ceiling lights in 1 2 3 4 5

my work area bothers me.
15. I would like to have an adjustable light on my desk. 1 2 3 4 5
16. Overall, I am satisfied with my workspace and furniture. 1 2 3 4 5
17. My space is poorly lighted. 1 2 3 4 5
18. I am satisfied with the quality of my work. 1 2 3 4 5
19. The lighting in my office is too dim. 1 2 3 4 5
20. The amount of light in my office washes out 1 2 3 4 5

the computer display.
21. My office is humid. 1 2 3 4 5
22. I am able to easily adjust my viewing distance 1 2 3 4 5

to the computer screen.
23. The overhead lights create distracting reflections 1 2 3 4 5

in my computer screen.
24. My computer screen has an anti-glare device. 1 2 3 4 5
25. I can easily adjust the brightness of my computer display. 2 3 4 5
26. My office is cold. 2 3 4 5
50
27. The lighting in my office is too bright. 1 2 3 4 5
28. The lights in my work area make a loud 1 2 3 4 5

humming/buzzing sound.
29. My space is stimulating. 1 2 3 4 5
30. I am satisfied with the lighting in my work space 1 2 3 4 5
31. My office is hot. 1 2 3 4 5
32. The work atmosphere in my office is tense. 1 2 3 4 5
33. The space is adequate for my job. 1 2 3 4 5
34. My space appears spacious. 1 2 3 4 5
Part B. Instructions: Please evaluate the following items by circling the appropriate
number.
Coding Scheme:
1 Excellent
2 Good
3 Adequate
4 Fair
5 Poor
1. Amount of space for your work. 1 2 3 4 5
2. The quality of the lighting in the corridors and hallways. 1 2 3 4 5
3. Condition of desks and chairs. 1 2 3 4 5
4. Noisiness of the space. 1 2 3 4 5
5. Amount of lighting. 1 2 3 4 5
6. Color of furniture and objects. 1 2 3 4 5
51
7. The quality of the lighting in the conference rooms. 1 2 3 4 5
8. Conversational privacy. 1 2 3 4 5
9. Maintenance of the space. 1 2 3 4 5
10. Appearance of the lighting system. 1 2 3 4 5
11. Ability to adjust light for work. 1 2 3 4 5
12. The quality of the lighting in my work space. 1 2 3 4 5
13. Ventilation and air circulation. 1 2 3 4 5
14. Pleasantness of the space. 1 2 3 4 5
15. Room temperature. 1 2 3 4 5
16. Cleanliness of the space. 1 2 3 4 5
17. The quality of the lighting in the restrooms. 1 2 3 4 5
18. The appearance of the work spaces and offices. 1 2 3 4 5
Follow-up:
Did you complete this same survey in [Insert date of Yes No
questionnaire distribution prior to modifications.
This question should only be included on the
post-modification survey]?
On the average, how many hours during the day are the lights in your office turned on?
0-2 2-4 4-6 6-8 8+
On the average, how many hours per day do you work at a personal computer?
0-2 2-4 4-6 6-8 8+
On the average, how many hours per day do you spend reading and writing paper tasks?
0-2 24 4-6 6-8 8+
52
Part C. For this portion of the survey, please record your response directly on these
sheets by placing an X on the line preceding your choice Unless indicated
otherwise, mark only one choice.
Cl. How long have you worked in your current work space?
A. Less than 1 year

B. 1-2 years
C. 2-5 years
D. More than 5 years
C2. The window in my office faces
A. North
B. South
C. East
D. West
E. I don’t have a window.
C3. Which of the following best describes
A. Administrative
B. Clerical
C. Technical
D. Other. Please specify:
C4. Do you
A. Wear single vision glasses?

B. Wear Bifocals or Trifocals?
C. Wear Contact Lenses?
D. None of the Above.
C5. How old are you?
A. less than 20
B. 20-29
C. 30-39
D. 40-49
E. 50-59
F. 60 or over
53
C6. Suppose you could make 4 changes to your overall work environment. Using the
list below, indicate the 4 changes you would make in order of preference (where 1 =
most preferred.
_ A. More comfortable day-to-day temperatures

_ B. More privacy
_ C. Change in color of furnishings or carpet
_ D. Improved lighting
__ E. Less noise
_ F. Improved air circulation
_____ G. More surface area for work
_ H. More comfortable furnishings
_ I. More frequent cleaning
_ J. Adjustable task lighting
_ K. More adjustable chair
_ L. View out/daylight
_ M. Other
C7. Please explain the reasons for the four choices you made in question C6.
Choice 1:_
Choice 2:_
Choice 3:_
Choice 4: __
This question was added for the second questionnaire
C8. How does the newly installed lighting in your office compare to the old lighting?
Thank you very much for your participation!

Appendix B. Suggested Physical Measurement Form
Location Number:.
Today’s Date:_
Current Time:_
DOE/POE
Environmental Measures
Room Inventory
A. Workstation Dimensions Length Width Height
1. Office Area _ _ _
2. Desk _ _
3. File Length _
B. Luminaires Active Inactive Total
1. Number of Luminaires/Work Space _ _ _
2. Number of Lamps/Luminaire _ __
3. Number of Lamps/Work Space _ _ _
4. Control/Switching None On/Off
C. Photometric Measurements
WITH TASK LIGHT ON
1. Worksurface Illuminance (lx): Mean Illuminance_
Far Left Far Right
Near Left Near Right
55
Location Number:
2. Task Luminance (cd/m2)
Gloss Black _ Matte Black
Gloss White _ Matte White
3. Room Surface Illuminances (lux):
90 degrees 45 degrees Straight 45 degrees 90 degrees

to the Left to the Left Ahead to the Right to the
Right
2 m AFF*
1 m AFF
* Above Finis led Floor
4. F.oom Surface Luminances (cd/m2):

Right
2 m AFF*
1 m AFF
"Above Finis ted Floor
5 Ceiling: Brightest (Furthest Luminaire): cd/m2

Darkest: _ cd/m2
56
Location Number:
D. Supplemental Task Lighting

1. Which task location? Primary Secondary None
2. Lamp Type_
3. Mounting:
Free Standing/Movable Y N
Under Cabinet Y N
Desk Mounted/Movable Y N
Other_
4. Height Above Task Surface_cm
E. Photometric Measurements
WITH TASK LIGHT OFF
1. Worksurface Illuminance (lx): Mean Illuminance
Far Left Far Right
Near Left Near Right
2. Task Luminance (cd/m2)
Gloss White _____ Matte White
57
Location Number:
3. Room Surface Illuminances (lux):

Right
2 m AFF*
1 m AFF
* Above Finis ted Floor
4. Fx)om Surface Luminances (cd/m2):

Right
2 m AFF5"
1 m AFF
"'Above Finished Floor
5. Ceiling: Brightest (Furthest Luminaire):_cd/W

Darkest:_cd/m2
F. Windows
1. Is there a window? Y N
2. Orientation of the window: N E S W
3. Does the task "see" the window? Y N
4. Ratio of Window Area/Wall Area %
5. Exposure ratios Sky _% cd/m2

and Luminances Buildings _% cd/m2
6. Shading Devices Used? _ Shades _ Curtains Venetian

Blinds
Portion of window occluded.
58
Location Number:
G. VDTs
1. Does the workstation have a VDT? Y N
2. Can reflections of electric light

sources be seen on the screen? Y N
3. Is screen position adjustable? Y N
4. Proximity of Window to SCREEN Behind Front Left
5. Check source document location
6. Source Document Photometry

a. Task Illuminance (lx)_
b. Luminance (cd/m2) Task Light On
59
Location Number:
H. Workstation Materials
1. Dominant workstation color R O Y G Bu P Bk Br
Wh
Chromaticity coordinates
x_ y_ Y_ -r E_ = p_
2. Walls are predominantly

Fabric Y N
Masonry Y N
Wood Y N
Metal Y N
Dry Wall Y N
Demountable Ceiling Y N
Height Partitions Y N
Partial Height Partitions Y N
If yes, how tall? 42"-48" 60”-66"

48"-54" 66" -72"
4^
Ln
ON
o
i
Workstation is an open area? Y N
The space containing the

workstation is ]Private Shared (How many?
Furnishings:
Pictures or Posters on Wall Y N
Pictures/mementos on desk, Y N
Credenza, tack surface Y N
Fan Y N
Space Heater Y N
60
Location Number:
G. VDTs
1. Does the workstation have a VDT? Y N
2. Can reflections of electric light

sources be seen on the screen? Y N
3. Is screen position adjustable? Y N
4. Proximity of Window to SCREEN Behind Right Front Left
5. Check source document location
vertical
horizontal
6. Source Document Photometry

a. Task Illuminance (lx)_
b. Luminance (cd/m2) Task Light On
61
NIST-114 U.S. DEPARTMENT OF COMMERCE (ERB USE ONLY)
(REV. 6-93) NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY ERB CONTROL NUMBER DIVISION
ADMAN 4.09 210

PUBLICATION REPORT NUMBER CATEGORY CODE
MANUSCRIPT REVIEW AND APPROVAL
INSTRUCTIONS: ATTACH ORIGINAL OF THIS FORM TO ONE (1) COPY OF MANUSCRIPT AND SEND TO PUBLICATION DATE NUMBER PRINTED PAGES
THE SECRETARY, APPROPRIATE EDITORIAL REVIEW BOARD
TITLE AND SUBTITLE (CITE IN FULL)
Post-Occupancy Evaluation of the Forrestal Building
CONTRACT OR GRANT NUMBER TYPE OF REPORT AND/OR PERIOD COVERED
__NISTIR_
AUTHOR(S) (LAST NAME. FIRST INITIAL. SECOND INITIAL) PERFORMING ORGANIZATION (CHECK (X) ONE BOX)
NIST/GAITHERSBURG
Sanders, Philip A.
NIST/BOULDER
Collins, Belinda L.
JILA/BOULDER
LABORATORY AND DIVISION NAMES (FIRST NIST AUTHOR ONLY)
Building and Fire Research Laboratory, Building Environment Department

SPONSORING ORGANIZATION NAME AND COMPLETE ADDRESS (STREET, CITY. STATE. ZIP)

Washington, D.C.
PROPOSED FOR NIST PUBLICATION

JOURNAL OF RESEARCH (NIST JRES) MONOGRAPH (NIST MN) LETTER CIRCULAR
J. PHYS. & CHEM. REF. DATA (JPCRD) NATL. STD. REF. DATA SERIES (NIST NSRDS) BUILDING SCIENCE SERIES
HANDBOOK (NIST HB) FEDERAL INF. PROCESS. STDS. (NIST FIPS) PRODUCT STANDARDS
SPECIAL PUBLICATION (NIST SP) LIST OF PUBLICATIONS (NIST LP) OTHER
TECHNICAL NOTE (NIST TN) NIST INTERAGENCY/INTERNAL REPORT (NISTIR)

X
PROPOSED FOR NON-NIST PUBLICATION (CITE FULLY) X U.S. FOREIGN PUBLISHING MEDIUM
PAPER j_| CD-ROM
X
DISKETTE (SPECIFY)
OTHER (SPECIFY)
SUPPLEMENTARY NOTES
ABSTRACT (A 2000-CHARACTER OR LESS FACTUAL SUMMARY OF MOST SIGNIFICANT INFORMATION. IF DOCUMENT INCLUDES A SIGNIFICANT BIBLIOGRAPHY OR
LITERATURE SURVEY, CITE IT HERE. SPELL OUT ACRONYMS ON FIRST REFERENCE.) (CONTINUE ON SEPARATE PAGE, IF NECESSARY.)
A post-occupancy evaluation was performed on the Department of Energy Headquarters Building (the Forrestal Building) in
Washington, D.C. The lighting in the building was retrofitted with new, more energy-efficient, components to meet energy target
guidelines. Occupant responses to the indoor environmental conditions, particularly the lighting, were studied to determine the impact
of the relighting on the building inhabitants. In addition, physical measures of the lighting and other environmental conditions before
and after the relighting were compared. The post-occupancy evaluation employed a questionnaire about the environmental conditions
and physical measures of the space (lighting, space, noise, temperature, etc.). A total of 244/220 people participated (before and after
relighting, respectively). Physical measures were taken at 100 work stations before the relighting and 75 after. Analysis of the
physical measurement data indicated generally higher lighting levels with more even distribution of luminances in the offices.
Occupant response to the changes in the lighting was generally quite positive. The relighting was perceived to have improved the
appearance of the building substantially, as well as the lighting within individual workstations. Finally, guidance is given for doing
post-occupancy evaluations as part of other relighting initiatives.
KEY WORDS (MAXIMUM OF 9; 28 CHARACTERS AND SPACES EACH; SEPARATE WITH SEMICOLONS; ALPHABETIC ORDER; CAPITALIZE ONLY PROPER NAMES)
building technology; contrast; energy efficiency; environmental assessment; federal relighting; illuminance; lighting; luminance; post¬
occupancy evaluation; temperature; VDT’s
AVAILABILITY NOTE TO AUTHOR(S); IF YOU DO NOT WISH THIS
UNLIMITED | | FOR OFFICIAL DISTRIBUTION - DO NOT RELEASE TO NTIS MANUSCRIPT ANNOUNCED BEFORE PUBLICATION.
ORDER FROM SUPERINTENDENT OF DOCUMENTS, U.S. GPO. WASHINGTON. DC 20402 PLEASE CHECK HERE. n
ORDER cROM NTIS, SPRINGFIELD. VA 22161
WORDPERFECT

Ni Stir 5591

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ni Stir 5591

Uploaded by

Copyright:

Available Formats

NISTIR 5591

U.S. DEPARTMENT OF COMMERCE

U.S. Department of Energy

U.S. DEPARTMENT OF COMMERCE

U.S. Department of Energy

U.S. DEPARTMENT OF COMMERCE

NATIONAL INSTITUTE OF STANDARDS

A post-occupancy evaluation was performed on the Department of Energy Headquarters Building

6. Recommendations for Completing Further Post Occupancy

Appendix A. Suggested Questionnaire Items. 49

Appendix B. Suggested Physical Measurement Form . 55

Figure 1. Ratings of the Quantity of Light as a Function of Sequence. 11

Table 1. Summary of Questionnaire Results for Sequence and Window

Post-occupancy evaluation techniques provide a means for evaluating occupant responses to

Post-occupancy evaluation techniques, including questionnaires and physical measurements, were

2.2.1 Questionnaire Survey

2.2.2 Physical Measurements

Lb = luminance of the background, and

CIE stands for Commission Internationale de l’Eclairage or the International

North 29 (12 percent) 27 (12 percent)

3.2 Subjective Response to Lighting

Figure 4 Ratings of the General Amount of Space as a Function of Window

3.4 General Appearance of the Offices

3.5 Reasons for Choices

More adjustable chair

Adjustable task lighting

More frequent cleaning

More comfortable furnishings

More surface area

Improved air circulation

Strongly Somewhat Somewhat Strongly

A: Agree Agree Neutral Disagree Disagree

B: Excellent Good Adequate Fair Poor

Mean Mean Difference P

A The amount of light 2.11 1.73 0.38 0.00 Better

A The ceiling lights in 3.26 3.54 -0.29 0.02 Less

A The light in my office 3.08 2.73 0.35 0.00 Better

A The irregular pattern 3.26 3.76 -0.50 0.00 Less

A My space is poorly 3.51 4.03 -0.52 0.00 Better

A The lighting in my 3.76 4.10 -0.34 0.00 Brighter

The quality of the 3.47 3.10 0.37 0.00 Better after

The quality of the light 3.14 2.79 0.35 0.00 Better

East None West South North

General disagreement implying not enough light.

A I would like to have an adjustable light on my desk

South North None East West

General agreement although East and West exposures less.

None East North South West

General agreement with those having windows agreeing less

South West North East None

Generally agreeable (None - windowless - less agreeable)

B Amount of space for your work

East West North South None

Generally Good, with Windowless Adequate to Fair

A The air circulation in my office is adequate

West East North South None

General disagreement. South and None disagree more strongly.

None South North East West

Generally disagree though S and None were more agreeable

None South East North West

Generally agreeable especially for South and None

B Ventilation and air circulation

West East North South None

A The overall appearance of offices is pleasing